1. 1. An apparatus is described for analysing the patterns of sounds produced by termites. 2. 2. The convulsive movement ( Vertical Oscillatory Movement) of the soldier Z. angusticollis results in the production of sound when the head comes into contact with the roof and floor of a container approximately the same height as a nest-gallery. The same movement in the larvae and nymphs results in the head being hit against the roof only. 3. 3. The activity of the termites is found to be affected by substratum vibration but not by airborne sound. Under the influence of substratum vibration, positive geotaxis, negative phototaxis, positive “chemotaxis”, and positive “thigmotaxis” in the termites are initiated, or become more marked. 4. 4. A group of termites that have been kept on one particular area of a sloping surface move downwards from this area if the support is tapped. If the tapping is continued they will stay down, but if it is discontinued after about one minute they will gradually reaccumulate in the area they originally occupied. 5. 5. On the basis of the experimental evidence it is suggested that the substratum vibration produced by the termites has the same effect on them as a warning signal would be expected to have, dispersing them to the deeper and more inaccessible galleries. 6. 6. Once the termites have been subjected to a disturbance, contact between individuals is important in causing the release of the Vertical Oscillatory Movement. Two necessary features of this contact stimulus are the presence of a species-specific pheromone, which is probably detected by sensilla on the antennae, and some degree of movement on the part of the object contacted. 7. 7. Substratum vibration produced by the termites themselves acts to some extent as a releaser for the Vertical Oscillatory Movement. 8. 8. Certain features of the pattern of sound produced by the termites have been found to be constant at a given temperature. These have been referred to as (i) the repetition period; (ii) the ratio of the numbers of 2- to 3-sound-groups, which is usually constant; (iii) the ratio of the numbers of even- to odd-number sound-groups; (iv) the time between the first and last sounds of sound-groups containing the same number of sounds; and (v) the times between successive sounds in sound-groups containing the same number of sounds. The factors (i), (ii), (iv), and (v) vary similarly with respect to temperature. 9. 9. In view of the sometimes remarkable constancy of features of the sound-pattern, it is suggested as likely that the mechanism of control of the Vertical Oscillatory Movement is confined to the central nervous system. 10. 10. A hypothetical neurone circuit is described, the mode of action of which could explain the observed characteristics of the pattern of sounds produced by the termites. 11. 11. The relation between the pattern of sounds and the electrophysiological findings on the properties of the termite subgenual organ is discussed. On the basis of this, it is argued that a “language” making use of variations in the sound pattern could not be expected in Z. angusticollis.